The effect of rumination on recall of emotional words: comparison of dysphoric individuals with and without a history of nonsuicidal self-injury

ABSTRACT

Prior research and theory has suggested that rumination plays a role in nonsuicidal self-injury (NSSI), and rumination increases recall of negative autobiographical information in dysphoric individuals. Across two studies, we investigated whether rumination (versus distraction) differentially influences the recall of emotional words among dysphoric persons with and without a history of NSSI. Participants encoded unpleasant, neutral, and pleasant words and then were randomly assigned to either focus on the meaning and consequences of their emotions (i.e. rumination) or unrelated thoughts (i.e. distraction) before they were asked to recall encoded words. Across the two studies, we did not find a significant effect of rumination on memory for emotional words among dysphoric people with (Studies 1 and 2) or without a history of NSSI (Study 1). We did find that people were more likely to remember neutral words as opposed to unpleasant or pleasant words across studies, regardless of rumination condition. Together, results from these two well-powered studies provide fairly compelling evidence that rumination after encoding has little to no effect on recall for emotional words in people elevated on symptoms of depression or with NSSI history. These findings can be used to refine theories of rumination and NSSI.

KEYWORDS:

• Nonsuicidal self-injury
• NSSI
• rumination
• memory
• recall

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 We note that this power analysis is for a different, simpler statistical model than we used (see below). After collection of the data, we used the simr package in R (Green & MacLeod, 2016) to determine power for effect sizes based on our sample size (165 Level 1 units, 48 Level 2 units) and estimates of variance. These simulations showed that we had .96 power for OR = 1.16 and .76 power for OR = 1.05. An OR = 1.2 is equivalent to a one word difference between conditions, which is the smallest meaningful difference between conditions. Thus, we had adequate power to test the smallest meaningful difference. Because these results are post hoc and rely on estimates from our data, they should be interpreted with caution (Green & MacLeod, 2016).

2 Given our a priori interest in the NSSI group effects, we probed further by examining the word contrast by condition interactions within NSSI status. For individuals without a history of NSSI, neither contrast interacted with condition (p’s = .561 and .935). For individuals with a history of NSSI, there was a marginal interaction between condition and the contrast between neutral and unpleasant words, γ = .10, t(3498) = 1.92, p = .055, OR = 1.11, 95% CI [.99, 1.24]. This interaction showed that NSSI group participants in the rumination condition remembered more neutral words (M = .27, 95% CI [.23, .32]) in comparison to unpleasant words (M = .18, 95% CI [.14, .22]) than NSSI participants in the distraction condition (Neutral words: M = .27, 95% CI [.21, .32]; unpleasant words: M = .24, 95% CI [.19, .29]), with the specific difference being a reduction in the number of unpleasant words recalled in the rumination condition. Because the interaction was not significant, we viewed these results with caution.

Additional information

Funding

Support for this work came from the National Institute on Drug Abuse Grant F31 DA038417 awarded to Konrad Bresin